Year: 2025
Type: article
Abstract: This paper presents the importance and applicability of Dissolved Gas Analysis (DGA) for monitoring the health of power transformers, essential in electrical distribution networks. The reliability of transformers is critical, and DGA represents an effective method for early detection of internal faults by measuring dissolved gases in the insulating oil. Different interpretation techniques, such as the Duval Triangle and Roger methods, are discussed for identifying fault types based on the concentration and ratio of the resulting gases. The paper also explores the limitations of these methods and future perspectives, proposing a multi-method approach based on machine learning algorithms to improve the accuracy and consistency of DGA interpretation.
Authors: Universitatea Natională de Știință si Tehnologie Politehnica București, Dan - Gabriel RADU, Valentin NĂVRĂPESCU, Universitatea Natională de Știință si Tehnologie Politehnica București
Language: English
Subfield: Electrical and Electronic Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4407667299